As a result of an extensive before-and-after accident study, the high-accident location, spot-improvement program in Kentucky, although not a costly program, was found to have significantly decreased the number of motor vehicle accidents. Favorable benefit-cost ratios indicated that the cost of the program has represented a good investment in comparison with the resultant savings in accident costs. The spot-improvement program had little effect on average accident severity as measured by a severity index. Detailed analysis of available accident data showed that, for studies of the type reported, the 12-month period immediately prior to the date of identification of a high-accident location is not a reliable period for representing the actual long-term "before" accident experience. A much more acceptable period is the 12 months beginning 2 years in advance of the date of identification. Further analysis also showed that a route segment of 0.1 mile (0.16 km) is not of sufficient length for properly identifying high-accident locations or for accumulating accident statistics to support a before-and-after study. Since only slightly more than five percent of the identified high-accident locations were judged to warrant improvement, the procedure used in Kentucky for identifying high-accident locations, namely, those having three or more accidents at a 0.1-mile (0.16-km) location during a 12-month period, was found to be inefficient.

Report Date


Report Number

No. 440

Digital Object Identifier



Accepted for publication by the American Society of Civil Engineers.